Author + information
- Received September 1, 2009
- Revision received October 23, 2009
- Accepted November 3, 2009
- Published online February 23, 2010.
- Pieter G. Postema, MD*,
- Pascal F.H.M. van Dessel, MD, PhD*,
- Jan A. Kors, PhD§,
- Andre C. Linnenbank, PhD‡∥,
- Gerard van Herpen, MD, PhD§,
- Henk J. Ritsema van Eck, MD, PhD§,
- Nan van Geloven, MSc*,†,
- Jacques M.T. de Bakker, PhD‡∥,
- Arthur A.M. Wilde, MD, PhD*,‡ and
- Hanno L. Tan, MD, PhD*,‡,* ()
- ↵*Reprint requests and correspondence:
Dr. Hanno L. Tan, Department of Cardiology, Academic Medical Center, Meibergdreef 9, 1105AZ Amsterdam, the Netherlands
Objectives We sought to obtain new insights into the pathophysiologic basis of Brugada syndrome (BrS) by studying changes in various electrocardiographic depolarization and/or repolarization variables that occurred with the development of the signature type 1 BrS electrocardiogram (ECG) during ajmaline provocation testing.
Background BrS is associated with sudden cardiac death. Its pathophysiologic basis, although unresolved, is believed to reside in abnormal cardiac depolarization or abnormal repolarization.
Methods Ajmaline provocation was performed in 269 patients suspected of having BrS with simultaneous recording of ECGs, vectorcardiograms, and 62-lead body surface potential maps.
Results A type 1 ECG was elicited in 91 patients (BrS patients), 162 patients had a negative test result (controls), and 16 patients had an abnormal test result. Depolarization abnormalities were more prominent in BrS patients and were mapped to the right ventricle (RV) by longer right precordial filtered QRS complex durations (142 ± 23 ms vs. 125 ± 14 ms, p < 0.01) and right terminal conduction delay (60 ± 11 ms vs. 53 ± 9 ms, p < 0.01). Repolarization abnormalities remained concordant with depolarization abnormalities as indicated by steady low nondipolar content (12 ± 8% vs. 8 ± 4%, p = NS), lower spatial QRS-T integrals (33 ± 12 mV·ms vs. 40 ± 16 mV·ms, p < 0.05), similar spatial QRS-T angles (92 ± 39° vs. 87 ± 31°, p = NS), similar Tpeak-Tendinterval (143 ± 36 ms vs. 138 ± 25 ms, p = NS), and similar Tpeak-Tenddispersion (47 ± 37 ms vs. 45 ± 27 ms, p = NS).
Conclusions The type 1 BrS ECG is characterized predominantly by localized depolarization abnormalities, notably (terminal) conduction delay in the RV, as assessed with complementary noninvasive electrocardiographic techniques. We could not define a separate role for repolarization abnormalities but suggest that the typical signs of repolarization derangements seen on the ECG are secondary to these depolarization abnormalities.
- body surface potential mapping
- Brugada syndrome
- sudden cardiac death
Brugada syndrome (BrS) is associated with sudden cardiac death by ventricular tachycardia and/or fibrillation (VT/VF) (1). The ongoing controversy about the best treatment (2) requires better understanding of its pathophysiology. The signature type 1 electrocardiogram (ECG), mandatory for the BrS diagnosis and linked to VT/VF, consists of right precordial coved-type ST-segment elevation. In BrS patients with a type 2, type 3, or normal ECG at baseline, provocation testing with sodium channel–blocking drugs (e.g., ajmaline) is used to evoke a type 1 ECG (1). Because electrocardiographic abnormalities are found in right precordial leads and VT/VF mostly originates from the right ventricle (RV), BrS is regarded an RV disorder (1). There are 2 principal hypotheses on its pathophysiologic basis (reviewed in Meregalli et al. ): the “depolarization disorder hypothesis,” that is, RV conduction delay combined with mild structural RV derangements (4–6), and the “repolarization disorder hypothesis,” that is, transmural dispersion of RV action potential morphology (7). We aimed to gain more pathophysiologic insight by using 3 noninvasive electrocardiographic assays (ECG, vectorcardiogram [VCG], and body surface potential map [BSPM]) in parallel in patients who underwent ajmaline testing. This allowed us to analyze which changes occurred when a type 1 ECG emerged. Using the synergy of the different sets of information on global and local depolarization and repolarization events from these different assays, we found that local depolarization abnormalities are the dominant pathophysiologic mechanism of BrS.
Patients and the ajmaline protocol
In this study, 269 consecutive subjects who underwent ajmaline testing were included. These subjects had not previously exhibited a type 1 ECG and underwent testing because of symptoms, a suspicious ECG, or family screening. The type 1 BrS ECG was defined as ≥0.2 mV J amplitude with coved ST-T segments gradually descending into a negative T-wave in ≥2 right precordial leads (1). Studies, performed after written informed consent, were approved by the institutional ethics committee.
Ajmaline was infused in repeated boluses of 10 mg/min until the target dose (1 mg/kg) was reached, and a type 1 ECG or arrhythmias occurred. Test results were defined as positive if ajmaline infusion evoked a type 1 ECG (and subjects as BrS patients when the other diagnostic criteria were also met ). Tests in which excessive QRS widening (>40% of baseline values) or arrhythmias occurred, but no type 1 ECG, were defined as abnormal and were excluded from further analysis. All other tests were defined as negative (and subjects as controls). In BrS patients or those with an abnormal test result, DNA analysis for SCN5a(the gene that encodes the α-subunit of the cardiac sodium channel) mutations was performed. Because SCN5amutations and sex may modify disease expressivity (1), secondary analyses on these modifiers were performed.
In all 269 subjects, ECGs were recorded as described previously (8). From modified 12-lead ECGs (conventional V3and V5positioned in the third intercostal space cranially to V1and V2[V1ic3 and V2ic3, respectively]) (Fig. 1),10-s recordings were saved at baseline, at 1 minute after each ajmaline bolus, and during wash out. VCGs were recorded simultaneously with ECGs in all subjects and additional Frank XYZ leads (E/I/M/H; C = V4, A = V6, F = left leg) (Fig. 1). Additional continuous BSPM recordings before, during, and after ajmaline provocation were made in the first 95 subjects with a customized recording system using a grid of 62 unipolar leads on 14 flexible straps (Fig. 1), as described previously (9–11).
ECG and VCG analyses
All digital electrocardiographic and vectorcardiographic recordings were first automatically analyzed with the modular ECG analysis system MEANS (12) and manually verified. To aid differentiation between depolarization and repolarization abnormalities at baseline and at peak ajmaline effect, we studied simultaneous recordings at these time points.
From the ECG, we analyzed conventional depolarization parameters (P, PQ, QRS [all in milliseconds] and the QRS axis [in degrees]) and repolarization parameters (JT, QT, QTc [all in milliseconds] and the T axis [in degrees]). T-wave end was defined with a tangent, and QT was corrected for heart rate (QTc) using Bazett's formula (13). The tallest right precordial J amplitude (μV) among V1, V2, V1ic3, and V2ic3 was determined (Fig. 2).As a measure of dispersion of repolarization (regarded as transmural or global dispersion of repolarization [14–16]), we studied maximal Tpeak-Tendinterval and Tpeak-Tenddispersion (maximal-minimal value) over the precordial leads. From this analysis, we excluded biphasic T waves and T waves with amplitudes <100 μV.
VCGs may reveal abnormalities in depolarization and/or repolarization sequence not available on a 12-lead ECG (Fig. 2) (17,18). To assess depolarization, we divided the spatial QRS loop into 4 parts of equal length. For each part, we measured the duration (in milliseconds) and axis (in degrees) in the transverse plane. To study interaction between repolarization and depolarization, we measured the spatial QRS T angle (in degrees) and the vector magnitude (mV·ms) of the spatial QRS T integral (19,20). Studying the transverse plane of the VCG allows assessment of RV terminal conduction slowing (17,18). Abnormally increased spatial QRS T angles implicate discordant repolarization and predict cardiovascular mortality (19). The vector magnitude of the spatial QRS T integral, or ventricular gradient, is a 3-dimensional measure of heterogeneity of action potential duration (20). It was postulated that increased transmural dispersion of RV action potential morphology in BrS (7) should result in an increased ventricular gradient (20).
A BSPM is better suited than a 12-lead ECG to uncover local abnormal electrophysiologic phenomena (9,11). BSPMs were processed using customized software (21). One-minute intervals were selected at baseline and peak ajmaline effect (similar to ECGs/VCGs). In this way, electrocardiograms could be averaged to exclude artifacts and reduce noise while preserving short-term effects during ajmaline infusion (ajmaline has a half-life of minutes). On average, per map 1.2 ± 0.3 leads were rejected because of unacceptable quality. Depolarization was assessed by late potential (LP) analysis and repolarization by nondipolar content. LPs indicate abnormal ventricular conduction and are related to re-entrant tachyarrhythmias (22). Nondipolar content permits evaluation of local ventricular repolarization abnormalities independent of ventricular activation sequence (23). High nondipolar content indicates local disparity of repolarization duration and has been used to assess vulnerability to ventricular arrhythmias (9–11,23).
For LPs, we assessed both global and local relevance. Global relevance was studied in reconstructed Frank leads and local relevance in sets of right and left precordial BSPM leads (Fig. 1). The right precordial set was used to evaluate the contribution of abnormal RV depolarization, and the left precordial set to assess this for the left ventricle. After application of a bidirectional Butterworth 25-Hz high-pass filter, the root mean square filtered signal was computed. Onset and end of the filtered QRS complex were taken at the time locations where signal amplitude became ≥3 μV for ≥5 ms. We applied this conservative and arbitrary fixed noise cutoff because we dealt with nonstationary signals between baseline and peak ajmaline effect. A noise-driven cutoff could be detrimental to between-subjects comparison (22). On average, noise levels were similar between BrS patients and controls and remained similar on ajmaline administration (Frank: 0.95 ± 0.35 μV, left precordial: 1.25 ± 0.46 μV, right precordial: 1.20 ± 0.39 μV). After verifying correct marker placement, we calculated the following parameters: the duration of the filtered QRS complex (fQRSD), the duration (in milliseconds) of the terminal low-amplitude signal <40 μV (LAS40), and the root mean square voltage (RMS40) (in microvolts) of signal in the terminal 40 ms) (Fig. 2).
To obtain nondipolar content, we calculated mean total QRS T integral maps as described previously (9,11). In short, the first 12 eigenvectors were derived, and the QRS T integral map was expressed in terms of these eigenvectors. The first 3 eigenvectors of the eigenvector set reveal (normal) dipolar patterns, whereas eigenvectors 4 through 12 show (increasingly abnormal) nondipolar or multipolar patterns (23). Therefore, the nondipolar content of each map as a measure of abnormal repolarization was defined as the contribution of eigenvectors 4 through 12 relative to the contribution of eigenvectors 1 through 12 and expressed as percentage (Fig. 2) (23).
Statistical analyses were performed using SPSS version 15.0 (SPSS Inc., Chicago, Illinois). Continuous variables are presented as mean ± SD and categorical variables as the number (percentage) of patients. Patients with an abnormal test result (see the “Patients and the ajmaline protocol” section) were excluded from comparisons. We studied the differences between BrS patients and controls at baseline and at peak ajmaline effect. Secondary analyses were performed to assess differences between BrS patients regarding SCN5amutations, symptoms, and sex. For continuous parameters, when histogram analyses confirmed normal distribution, an unpaired ttest was used to compare groups, with equal or unequal variances assumed according to Levene's test. For skewed distributions, a Mann-Whitney Utest was used. The Fisher exact test was used to compare categorical variables. Furthermore, based on LP analysis results, the presence of a correlation between right-to-left differences in fQRSD and maximal J amplitude in right precordial leads was estimated using Spearman's correlation coefficient. A p value <0.05 was accepted as the level of statistical significance.
Ajmaline test results were positive in 91 (34%), negative in 162 (60%), and abnormal in 16 (6%) subjects (Table 1).In 1 patient, sinus arrest occurred. All patients with a positive test result met the BrS consensus criteria (1). Age and sex distributions were similar, as were indications for testing and medical histories (Table 1). SCN5amutations were found in 20% of BrS patients (11 males, 7 females) and 44% of patients with an abnormal test result (3 males, 4 females).
Because ajmaline infusion was stopped when a type 1 ECG occurred, BrS patients received less ajmaline than controls (90 ± 19% vs. 102 ± 4% of the target dose, p < 0.01). BrS patients with SCN5amutations required less ajmaline than BrS patients without SCN5amutations (81 ± 22% vs. 92 ± 18% of the target dose, p < 0.05). Likewise, male BrS patients required less ajmaline than female BrS patients (85 ± 23% vs. 94 ± 14% of the target dose, p < 0.05). Symptomatic and asymptomatic BrS patients required similar amounts of ajmaline.
ECG depolarization and repolarization parameters
ECG parameters are summarized in Table 2.At baseline, BrS patients already had depolarization abnormalities with longer P, PQ, and QRS than controls (118 ± 16 ms vs. 114 ± 13 ms, 167 ± 30 ms vs. 156 ± 21 ms, and 108 ± 18 ms vs. 102 ± 11 ms, respectively; all p values <0.05). Ajmaline caused faster heart rates and atrial, atrioventricular, and ventricular conduction slowing in both groups. Although BrS patients received less ajmaline, more conduction slowing developed in them on ajmaline infusion, as evidenced by more widening of P waves (35 ± 17 ms vs. 30 ± 14 ms net increase, p < 0.01) and QRS duration (39 ± 20 ms vs. 31 ± 12 ms net increase, p < 0.01). Conversely, BrS patients had shorter JT intervals at baseline and peak ajmaline effect (266 ± 34 ms vs. 277 ± 28 ms and 254 ± 27 ms vs. 267 ± 27 ms, respectively; all p values <0.01), resulting in similar QT and QTc intervals in both groups. Maximal Tpeak-Tendintervals and Tpeak-Tenddispersion were also similar. BrS patients had more leftward deviation of the QRS axis and T axis, particularly at peak ajmaline effect (QRS axis: 13 ± 68° vs. 31 ± 62°, T axis: 31 ± 25° vs. 43 ± 17°, both p values <0.05). Obviously, right precordial J amplitude at peak ajmaline effect was greater in BrS patients (352 ± 161 μV vs. 141 ± 86 μV, p < 0.01) (Figs. 2 and 3,⇓Table 2).
Depolarization abnormalities on BSPM
LP analysis (Fig. 2, Table 3)revealed significant depolarization abnormalities in BrS patients on type 1 ECG induction. Global LP analysis, conducted with the use of Frank leads, revealed that BrS patients had lower RMS40 and longer LAS40, both at baseline and peak ajmaline effect (RMS40 baseline: 44 ± 23 μV vs. 62 ± 37 μV; RMS40 peak ajmaline effect: 15 ± 7 μV vs. 21 ± 15 μV; LAS40 baseline: 32 ± 11 ms vs. 28 ± 9 ms; LAS40 peak ajmaline effect: 52 ± 17 ms vs. 43 ± 12 ms; all p values <0.05). Furthermore, we found disparate results for the left ventricle and RV. In left precordial leads, LP measures were similar. In contrast, in right precordial leads, fQRSD and LAS40 were longer and RMS40 was lower at baseline in BrS patients. These differences became even greater at peak ajmaline effect (fQRSD: 142 ± 23 ms vs. 125 ± 14 ms, RMS40: 15 ± 9 μV vs. 28 ± 19 μV; LAS40: 51 ± 18 ms vs. 36 ± 15 ms; all p values <0.01) (Figs. 2 and 3, Table 3).
Depolarization abnormalities on VCG
VCG analysis (Table 4)revealed that most conduction delay in the spatial QRS loop resided in its terminal quarter in both patient groups. This was most apparent in BrS patients as the difference between BrS patients and controls increased even further at peak ajmaline effect (60 ± 11 ms vs. 53 ± 9 ms, p < 0.01) (Fig. 2, Table 4). The rightward deviation of the axis of this terminal quarter of the QRS loop when a type 1 ECG occurred (55 ± 39° vs. 69 ± 25°, p < 0.01) (Fig. 2, Table 4) indicated RV terminal conduction slowing (Fig. 2).
Repolarization abnormalities on BSPM and VCG
BSPM analysis did not reveal local repolarization abnormalities in BrS patients. Nondipolar content remained low and was similar to that of controls (baseline: 10 ± 5% vs. 8 ± 4%, peak ajmaline effect: 12 ± 8% vs. 10 ± 6%; p = NS) (Fig. 2, Table 3). Only a few subjects had abnormal nondipolar content of >19% (23) (baseline: 3 [8%] vs. 2 [4%], peak ajmaline effect: 6 [15%] vs. 4 [7%]; p = NS).
The ECG and BSPM finding that the occurrence of a type 1 ECG in BrS patients was not associated with local repolarization abnormalities was supported by VCG analysis (Table 4). BrS patients had lower ventricular gradients (baseline: 36 ± 14 mV·ms vs. 42 ± 18 mV·ms, peak ajmaline effect: 33 ± 12 mV·ms vs. 40 ± 16 mV·ms; both p values <0.05). Furthermore, the absence of discordant repolarization was apparent from normal spatial QRS T angles at baseline and peak ajmaline effect, regardless of whether a type 1 ECG occurred (peak ajmaline effect: 92 ± 39° vs. 87 ± 31°, p = NS) (Table 4).
Correlation between J amplitude and depolarization abnormalities
As final supportive evidence of the notion that BrS is associated with RV conduction disorders, we found a positive correlation between right precordial J amplitude and RV conduction-slowing indices. J amplitude correlated most closely with right-to-left difference in fQRSD (fQRSD from right precordial leads minus fQRSD from left precordial leads; Spearman's rho 0.411, p < 0.01) (Fig. 3C). Accordingly, BrS patients had higher J amplitudes and greater right-to-left fQRSD differences (352 ± 161 μV vs. 141 ± 86 μV and 25 ± 16 ms vs. 8 ± 10 ms, respectively; both p values <0.01) (Online Table S1). This correlation also explained the lack of differences in J amplitude between BrS patients with and without SCN5amutations, although fQRSD in the right precordial leads was greater in those with SCN5amutations (158 ± 23 ms vs. 135 ± 19 ms; p < 0.01) (Supplemental Table S2). Because, in BrS patients with SCN5amutations, the left precordial fQRSD was increased to a similar extent (132 ± 23 ms vs. 110 ± 17 ms; p < 0.01), the net right-to-left fQRSD difference between BrS patients with SCN5amutations and those without was similar (26 ± 16 ms vs. 25 ± 17 ms; p = NS); thus, J amplitudes were also similar (360 ± 167 μV vs. 351 ± 161 μV; p = NS).
Secondary analyses: SCN5a, sex, and symptoms
BrS patients with SCN5amutations had more depolarization disorders at baseline than those without (Supplemental Table S2). Although BrS patients with SCN5amutations received less ajmaline, the differences with regard to conduction velocity at atrial, atrioventricular, and ventricular levels remained of similar magnitude at peak ajmaline effect (P: 166 ± 18 ms vs. 153 ± 23 ms; PQ: 234 ± 39 ms vs. 212 ± 27 ms; QRS: 163 ± 31 ms vs. 143 ± 18 ms; duration of fourth quarter QRS loop: 66 ± 14 ms vs. 59 ± 10 ms; all p values <0.05). Differences with regard to all LP parameters in Frank and left and right precordial leads became even greater at peak ajmaline effect (fQRSD: Frank, 137 ± 25 ms vs. 115 ± 17 ms; left precordial, 132 ± 23 ms vs. 110 ± 17 ms; right precordial, 158 ± 23 ms vs. 135 ± 19 ms; all p values <0.01). In contrast, repolarization parameters on ECG, VCG, and BSPM were similar.
Depolarization differences between male and female BrS patients were somewhat smaller (Supplemental Table S3). Male BrS patients had more atrial and ventricular conduction slowing and higher fQRSD at baseline (P: 122 ± 15 ms vs. 115 ± 17 ms; QRS: 115 ± 19 ms vs. 102 ± 15 ms; duration of fourth quarter QRS loop: 50 ± 13 ms vs. 43 ± 11 ms; fQRSD: Frank, 101 ± 16 ms vs. 87 ± 8 ms; left precordial, 99 ± 19 ms vs. 85 ± 8 ms; right precordial, 114 ± 20 ms vs. 95 ± 10 ms; all p values <0.05). However, these differences were attenuated at peak ajmaline effect (note that female subjects received relatively more ajmaline). In addition, female subjects had longer QTc intervals (baseline: 406 ± 32 ms vs. 393 ± 26 ms, peak ajmaline effect: 472 ± 29 ms vs. 452 ± 24 ms; both p < 0.05) and lower ventricular gradients (baseline: 32 ± 13 mV·ms vs. 41 ± 15 mV·ms; peak ajmaline effect: 29 ± 11 mV·ms vs. 37 ± 12 mV·ms; both p values <0.01).
Depolarization differences between symptomatic (VT/VF, unexplained syncope) and asymptomatic BrS patients primarily showed a trend toward more (RV) conduction slowing (Supplemental Table S4). Most differences, existing at baseline (QRS duration: 117 ± 26 ms vs. 106 ± 16 ms; duration of fourth quarter QRS loop: 53 ± 18 ms vs. 45 ± 11 ms; right precordial fQRSD: 117 ± 31 vs. 102 ± 14; all p values <0.05), lost significance after addition of similar amounts of ajmaline (duration of fourth quarter QRS loop: 66 ± 13 ms vs. 59 ± 11 ms, p < 0.05). Nondipolar content at peak ajmaline was even lower for symptomatic BrS patients (8 ± 2% vs. 13 ± 8%, p < 0.05).
We used the synergy of simultaneously recorded ECG, VCG, and BSPM to study both global and local abnormalities of depolarization and repolarization during ajmaline provocation testing with the aim to better understand the pathophysiologic basis of BrS. The most distinctive findings associated with the development of the type 1 ECG were: 1) right terminal conduction delay in combination with 2) increase in LPs, which were localized, in particular, in right precordial leads, and 3) the absence of major global or local repolarization abnormalities.
With the ECG analyses, we confirm previous reports in which BrS patients showed excessive conduction slowing on sodium channel blockade at the atrial, atrioventricular, and ventricular level (8). We did not find independent repolarization abnormalities: QT and QTc interval and Tpeak-Tendinterval measures were similar between BrS patients and controls. Instead, repolarization adapted to conduction slowing with JT interval shortening.
Vectorcardiography emerged as a suitable method to quantify the characteristic RV conduction slowing associated with the type 1 ECG (4,6,18). Moreover, VCG allowed detailed analysis of repolarization by providing information on depolarization-repolarization concordance (spatial QRS T angle) and heterogeneity of action potential duration (ventricular gradient) (19,20). These analyses revealed that, although intense depolarization abnormalities occurred with the type 1 ECG occurrence, repolarization remained concordant. This suggests that the abnormal deep negative T waves in right precordial leads, which constitute the type 1 ECG, reflect an abnormal repolarization pattern that simply follows an abnormal depolarization pattern and that abnormal repolarization is not the inciting mechanism. Furthermore, instead of an increase in the ventricular gradient that would be associated with increased heterogeneity of action potential duration (20), the ventricular gradient decreased. This is consistent with the notion that the net left ventricular contribution to the QRS T integral is effectively cancelled by enhanced terminal RV conduction delay, a process previously observed in patients with pulmonary arterial hypertension (24).
BSPM analysis provided additional information on local electrophysiologic phenomena associated with the type 1 ECG. In particular, LP analysis of right precordial leads underscored the role of RV LPs in the type 1 ECG development (25). LPs can be caused by lengthening of the pathway of excitation and/or by conduction slowing, with structural disruptions as possible critical determinants (22). As impulse propagation is determined by both active and passive membrane properties, it is likely that LPs in BrS are accentuated by disruptions in both properties (6). Abnormal membrane excitability causing conduction slowing in BrS may follow from reduced sodium current through reduced sarcolemmal sodium channel density or changes in sodium channel function. The additive effect of ajmaline and SCN5amutations on LPs supports the role of decreased sodium channel availability. In addition, increased coupling resistance between cells (e.g., by separation of myocardial fibers by fibrosis or fat or by decreased connexin expression) will slow conduction, lengthen the excitation pathway, and result in a discontinuously traveling electrical impulse and LPs (4,6,22). Interestingly, fibrosis and reduced connexin expression may also result from decreased sodium current (4). Further, in accordance with the ECG and VCG analyses and earlier studies (4,5,25), repolarization on BSPM (nondipolar content, a sensitive marker of local disparity of repolarization duration ) did not indicate local repolarization abnormalities associated with the type 1 ECG or BrS. This further supports the notion that repolarization abnormalities follow the depolarization abnormalities because this results in a consistent QRS T integral over the complete body surface.
From the BSPM analyses, we found that right-to-left differences in conduction slowing are important. At the end of the QRS complex, these differences will constitute an electrical gradient above the right ventricular outflow tract in apicobasal and posteroanterior direction that may be of sufficient magnitude to inscribe as right precordial J elevation on the ECG. Although SCN5amutations are associated with more conduction slowing, they are not limited to the RV and, accordingly, do not result in greater J amplitudes.
It is readily conceivable how excessive RV conduction slowing, despite an apparent absence of distinct repolarization abnormalities as found in the present study, may induce arrhythmias. It is well known that the spontaneous or drug-induced appearance of the type-1 ECG in BrS is associated with a higher risk for arrhythmias and/or sudden cardiac death (1,26). Interestingly, the presence of (loss of function) SCN5amutations, although associated with more conduction delay, is not an independent risk factor (2). We found that RV, as opposed to left ventricular, conduction delay is important to create the type 1 ECG. This might well explain why SCN5amutations do not result in higher arrhythmic risk. The mechanism by which local conduction slowing results in an arrhythmic substrate probably lies in a combination of factors (6): local reduction of wavelength in combination with augmented anisotropy and discontinuous conduction due to interstitial changes and/or electrical uncoupling. This will promote conduction block, reentry, and wave break and may result in VT/VF. However, the prognostic value of larger right-left differences in conduction slowing, measured either invasively or noninvasively, needs to be confirmed in large and long-term follow-up studies.
In summary, by simultaneously recording ECGs, VCGs, and BSPMs during ajmaline provocation testing, we found that RV conduction slowing is essential in the development of the type 1 ECG in BrS. These findings support previous studies in which dominant depolarization abnormalities and no distinctive repolarization abnormalities were found (4–6,25). These depolarization abnormalities, rather than abnormal repolarization, seem to be the primary pathophysiologic basis of the coved-type ST segment with J elevation and deep negative T-wave.
First, although we found no evidence of independent repolarization abnormalities associated with the type 1 ECG, we cannot rule out that transmural dispersion of action potential morphology (7) plays a role. Although the techniques used are considered reliable methods to study abnormal cardiac electrophysiology and suitable to study (transmural) dispersion of action potential morphology (9,16,20,23), it is possible that they cannot detect subtle disparities. Second, we used various parameters to distinguish between global and/or local depolarization and repolarization abnormalities associated with the type 1 ECG. In view of the explorative nature of this study, correction for multiple testing was not performed (27). This could make our results more prone to type I errors. Because many of the studied parameters exhibited considerable overlap (as expected) and concordant presence or absence of significant differences was found, we are confident that this limitation is subsidiary. Finally, the secondary analyses into the role of SCN5amutations, sex, and symptoms were hampered by smaller numbers of subjects.
The type 1 BrS ECG is characterized predominantly by localized depolarization abnormalities, notably (terminal) conduction delay in the RV, as assessed with complementary noninvasive electrocardiographic techniques. We could not define a separate role for repolarization abnormalities but suggest that the typical signs of repolarization derangements seen on the ECG are secondary to these depolarization abnormalities.
The authors thank the patients for their cooperation and Dr. Andrés Pérez Riera (São Paulo, Brazil) for inspirational discussions.
For Supplementary Tables 1 to 4, please see the online version of this article.
This study was funded by the Netherlands Heart Foundation(grants 2005T024to Dr. Postema and 2005B092to Dr. Linnenbank); Fondation Leducq Trans-Atlantic Network of Excellence, Preventing Sudden Death(grant 05-CVD-01to Dr. Wilde); Royal Netherlands Academy of Arts and Sciences(to Dr. Tan); and the Netherlands Organization for Scientific Research(grant ZonMW-VICI 918.86.616to Dr. Tan).
- Abbreviations and Acronyms
- Brugada syndrome
- body surface potential map
- filtered QRS complex duration
- duration of low-amplitude signal (<40 μV) in the terminal part of the filtered QRS complex
- late potential
- root mean square value of voltage in the terminal 40 ms of the filtered QRS complex
- right ventricle/ventricular
- gene that encodes the α-subunit of the cardiac sodium channel
- ventricular tachycardia/ventricular fibrillation
- Received September 1, 2009.
- Revision received October 23, 2009.
- Accepted November 3, 2009.
- American College of Cardiology Foundation
- Paul M.,
- Gerss J.,
- Schulze-Bahr E.,
- et al.
- Meregalli P.G.,
- Wilde A.A.,
- Tan H.L.
- Coronel R.,
- Casini S.,
- Koopmann T.T.,
- et al.
- Kasanuki H.,
- Ohnishi S.,
- Ohtuka M.,
- et al.
- Postema P.G.,
- van Dessel P.F.,
- de Bakker J.M.,
- et al.
- Peeters H.A.,
- SippensGroenewegen A.,
- Wever E.F.,
- et al.
- Dambrink J.H.,
- SippensGroenewegen A.,
- van Gilst W.H.,
- et al.
- Perez Riera A.R.,
- Ferreira C.,
- Schapachnik E.
- Kardys I.,
- Kors J.A.,
- van der Meer I.,
- et al.
- Breithardt G.,
- Cain M.E.,
- el-Sherif N.,
- et al.
- Abildskov J.A.,
- Green L.S.,
- Lux R.L.
- Henkens I.R.,
- Mouchaers K.T.,
- Vonk-Noordegraaf A.,
- et al.